摘要
针对传统融合模型不能根据差异特征变化而动态调整造成融合图像效果差的问题,提出了红外偏振与光强图像拟态融合概念,论述了图像拟态融合的特性;研究了拟态感知机理和拟态变换定理,指出了图像拟态融合原理和模型的研究内容及实现途径;分析了融合算法类集的结构化和融合结构的可重构化两个关键技术问题.从而探索出通过改变融合模型结构满足不同模态图像动态变化需求的融合新理论、新方法.
The structure cannot be adjusted dynamically in the current fusion model under the change of different features among images, leading to low fusion effect. The concept of mimic fusion between infrared polarization and intensity image is proposed in this paper and the characteristics of mimic fusion are discussed. Then, the mimic perception mechanism and transformation theorem are studied, and the research contents and implementation ways of image mimic fusion are pointed out. Lastly, two key technologies on the structured implementation of fusion algorithm class-set and reconstruction of fusion structure are analyzed. The novel fusion theory and method are explored to meet the dynamic change demand of different modality images by changing fusion model structure.
作者
杨风暴
YANG Feng-bao(School of Information and Communication Engineering, North University of China, Taiyuan 030051, China)
出处
《中北大学学报（自然科学版）》
CAS
北大核心
2017年第1期1-8,共8页
Journal of North University of China(Natural Science Edition)
基金
国家自然科学基金资助项目(61672472)
关键词
红外偏振图像
拟态融合
拟态仿生学
结构重构
infrared polarization image
mimic fusion
mimic bionics
structure reconstruction
